NEW YORK (GenomeWeb) – The Medical Device Innovation Consortium (MDIC), a public-private non-profit organization aiming to bridge the gap between various stakeholders in the medical device market, is soliciting input from some of those stakeholders for the first of five specific initiatives under its Clinical Diagnostic Program (ClinicalDx) to expedite the regulatory process for clinical diagnostics.
Over the last few months, ClinicalDx's Somatic Reference Sample (SRS) working group has drafted a list of variants or other targets that would be covered in an ideal set of reference materials used to develop and validate next-generation sequencing-based cancer tests. The group is currently collecting public input on this list of variants through the first week of December.
Concurrent to the SRS initiative, ClinicalDx working groups have also been established for the Surrogate Samples, IVD Real World Evidence, IVD Clinical Evidence, and Fingerstick as the program seeks to improve the speed and curb the costs of the IVD regulatory process.
US Food and Drug Administration Center for Devices and Radiological Health Director Jeff Shuren, in conjunction with LifeScience Alley former President and CEO Dale Wahlstrom, former Medtronic Chairman and CEO Bill Hawkins helped establish the MDIC in 2012. The consortium now contains several governmental agencies, academic groups, medical device companies, contract research organizations, health IT companies, and smaller non-profit organizations.
ClinicalDx Program Director Carolyn Hiller explained that the members identify problems between the regulatory and commercial spaces and recommend pathways to solve the issues. However, she emphasized that as part of MDIC, "we do not look at policy, but instead at regulatory science" by creating "tools that will bring clarity for established regulation."
Hiller noted that MDIC established ClinicalDx in 2015 with a steering committee that included representatives from diagnostic partners including Roche, Becton Dickinson, Abbot, Exact Sciences, Illumina, Hologic, Sysmex, Genomic Health, and The Translational Genomics Research Institute.
As a non-profit, MDIC project funding comes from membership dues and project specific grants from the FDA, Hiller explained. "We [now] have members in the industry that are pre-revenue, who haven't sold anything yet, to giants in the industry."
Susan Alpert, an MDIC fellow and member of the ClinicalDx steering committee noted that "nobody owns the environment, as everybody is an equal in the room, and you really can talk through the challenges in the field.
Sometimes the regulator is a bit of a purist, and wants the tool to be perfect, while the industry will say 'it's impossible to do, since here are the limitations,' which then forces the groups to collaborate and develop novel yet consistent and accurate ways to solve challenges for each of the players, especially the patients," Alpert added.
Reference materials
According to a project overview from the consortium, NGS oncology assay developers currently create their own contrived samples and sample mixes for oncology validation studies because well-characterized samples and reference materials don't exist. The SRS initiative therefore aims to develop new reference samples and public datasets that researchers can use to develop and improve the accuracy, reliability, and transparency of these assays.
For the first phase of the project, which kicked off in June at the FDA, the Somatic Working Group performed a landscape and gap analysis to identify specific reference samples it should develop, said Hiller. The working group intends to identify, characterize, and develop strategies for maintaining supply and stability of reference samples over time.
Hiller noted that the team is now compiling results from the landscape analysis and will release a report on the study in January on the MDIC website.
Meanwhile, the steering committee is accepting input from public groups as to which specific sample variants to prioritize.
The team eventually wants to offer samples in a variety of forms — cells, DNA, RNA, formalin-fixed paraffin-embedded tissue — that represent the multitude of allele fractions.
After the public comment period ends, the team will hold a workshop in January to review public input on the variant selection. Hiller noted that the group will also consider whether to use in silico reference materials as an additional way to represent several variants, variant types, or difficult-to-acquire variants.
In addition, the group will need to establish analysis tools and quality metrics to assess the performance of chosen reference materials.
Depending on how successful the first phase is, in phase II the team will pilot the production of several important NGS reference samples for the oncology community. After completing the initial samples, the team will identify and implement process improvements as it scales development to produce samples representing variants identified in Phase 1.
According to Hiller, the MDIC will either directly lead phase II or hand it off to a private commercial partner affiliated with the project. However, the team will still oversee production and characterization of the prioritized NGS reference samples.
"Our primary concern is that we want these reference samples to be available to the largest pharmaceutical companies [and] the innovative researchers in smaller universities working on limited budgets," Hiller explained. "We want everyone to have access to our samples."
Hiller noted that using reference samples will streamline and potentially remove steps in the regulatory process for diagnostics companies, as test developers may be able to use the databases of samples in support or in lieu of traditional clinical studies.
'Make-believe sample'
In addition to SRS, the ClinicalDx steering committee has begun the Surrogate Sample Project, which Hiller explained originated because IVD companies struggle to obtain or retain clinical specimens to test their assays. The team therefore believes that using surrogate samples will "foster innovation" when clinical specimens are not available.
"That's the challenge; how do you create a make-believe sample that is an accurate representation of what you'd get from a very sick patient," Alpert explained. "What are the rules, and how do you do that accurately?"
The group therefore developed a glossary of terms associated with specimen types including surrogate samples, assessing current terms, and removing inconsistencies in scientific literature. Afterward, the team established a framework that described when surrogate samples can and cannot be adequately used to simulate clinical specimens during in vitro diagnostic test development.
"[Surrogate samples] were called all types of things, so we put in a clear, [standardized] definition for them, and outlined principles and considerations for designing surrogate samples," Hiller noted. "We also put in a hierarchy, starting with patient samples, and then if that didn't work, you went down the list in order of preference for using a surrogate sample."
According to Hiller, the team completed the framework process in less than two years, releasing the document to the Clinical and Laboratory Standards Institute The CLSI is now developing "timely standards recognized by laboratories, industry, accreditors, and regulators as the way to improve testing."
Now Hiller's group is creating educational materials that both regulators and industry representatives can use by applying the framework for surrogate samples in case studies. Hiller argued that by working together on the project, regulators and industry groups will have access to the same education material and "have a shared language" to help guide future decisions on surrogate samples.
Hard evidence
ClinicalDx is also looking to speed up innovation for IVD tests while retaining safety and efficacy. As part of the IVD Real World Evidence (RWE) initiative, a multi-disciplinary team will build a framework for using real-world evidence as a source for supporting regulatory and reimbursement decision-making for IVD tools.
According to Hiller, the team has analyzed present and historical real-world evidence, examining how the evidence has been used and where industry experts have encountered barriers incorporating the data. The group is now working with the FDA on a draft version of the framework, which Hiller said will be put out for public comment in the future before the final version is released
"We are seeing how to apply the data as evidence, as well as taking into account data bias, quality, and criteria, or other considerations that can be made," Hiller explained.
Overall, the team aims to expedite patient access to medical device technology while reducing the number of patients exposed to risks from clinical trial participation. For industry, the team aims to cut down time needed to develop clinical evidence while lowering costs for clinical evidence generation.
For an IVD test to enter the US commercial diagnostic market, companies must demonstrate that the tool achieves high levels of analytical validity and clinical validity. To obtain payor coverage and drive adoption, firms must also offer evidence of clinical utility. However, Hiller noted that most firms achieve these requirements sequentially, slowing down the process and wasting time pushing the product to market.
The ClinicalDx steering committee has therefore established the IVD Clinical Evidence initiative, where the group plans to create a white paper that IVD manufacturers can use to decide how to develop credible evidence for the three requirements.
"When developing the test, companies can use the [white] paper as a thought piece, so they can develop evidence for the [requirements] simultaneously, instead of sequentially," Hiller explained. "Time is absolutely money, so if [companies] can think about what they need at the start, [they can] save money and time taking these tests to patients."
While the IVD Clinical Evidence Initiative was one of the first projects the ClinicalDx began back in 2015, Hiller acknowledged that the team has been working through some logistical challenges in order to draft the white paper. While the group planned to deliver a 12- to 15-page concept paper that laid an outline for the framework for developing an IVD by September 2016, Hiller said that the white paper is now at least 40 pages.
Alpert explained that "the different players in this space — industry, regulators, patients, and payors — all have slightly different needs and understanding of how this could be done. The idea, of course, is … [that] data needs to be clear [enough] so companies know what is needed and the regulators and payors get what they want and need the first time."
When finished, the documents will have a general scheme of how medical test claims are determined along with potential alternate methods to obtain market entry, as well as how medical test impact clinical outcomes.
Stuck on fingerstick
Another project within the ClinicalDx Program that has stalled is the Fingerstick Initiative, which Hiller said will aim to develop and define analytical validity study designs for point-of-care devices that use finger capillary whole blood specimens.
According to Hiller, the ClinicalDx steering committee established the project because the volumes needed for existing study designs — milliliters of venous blood or serum — do not match the small volume of blood collected in a fingerstick — 250 to 500 microliters — and often remain stable over a relatively long period of time.
Hiller explained that Precision/Reproducibility and method comparision are the most problematic areas when working on study designs. Her team is working to add clarity to help firms with study designs that the FDA would find acceptable for these performance tests.
Hiller acknowledged the initiative has encountered challenges regarding interfering compounds during the collection process.
"The complexity lies in the challenge of redesigning analytical validity tests that were originally intended to use a test tube or more of blood to now use a sample for which you can only source [a] few drops at once and that should be tested soon after collection," Hiller noted. "The new study design has to show to the regulatory agency that the test is safe and effective for use with capillary blood while also not adding undue burden on the test developers to conduct their analytical and clinical validation studies."
Overall, Alpert believes that the MDIC has spent time getting representatives from different groups to sit "at the same table without the product in front of them, so nobody has a vested interest in the product." Rather, the group's goal is to establish a common understanding of what patients need in the product, what companies and payors can provide at a reasonable cost, and how to efficiently move the right products through the regulatory process.
Alpert acknowledged that "it's been a rewarding but challenging effort," and that opening the lines of communication among the groups will help minimize the time needed to push a product to market that will ultimately benefit the patient.